The Remote Sensing Oil Spill SAR Image Denoising Research Based On Wavelet Thresholding Optimization

This paper uses the method of wavelet transform to process remote sensing image of oil spills on the sea. This research work has provided new research ideas and methods to remote sensing image de-noising technology which has strong the oretical study of the meaning and application.Wavelet Transform in time domain and frequency domain at the same time has good and localized nature. It doesn't only make the structure and texture of images manifested in different resolution levels, but also has capability of detecting edges (local mutation). And thus, the using of wavelet transform in removing noises can extract and save the marginal information which plays a major role in the vision. In the topic, wavelet theory will be introduced to oil spills at sea remote sensing images (SAR image),comparing several traditional methods of wavelet de-noising threshold , putting forward its own threshold optimization De-noising, oil spills on the sea and remote sensing image processing to enhance the beneficial exploration and research, in an attempt to find a flexible, simple and practical approach.This paper studies the wavelet transform resolution and more time-frequency characteristics of local, in-depth analysis of the distribution of wavelet domain coefficients, proposes the two methods that are based on wavelet threshold optimization to eliminate the oil spill at sea SAR image noise remote sensing . And compared with the traditional processing approach, these algorithms not only effectively eliminate noise, but also maintain the image edge details advantage. The actual remote sensing image has been processed and tested analyzed, and the effectiveness of these algorithms has been certified.This paper summarizes the major works from the following two aspects:1. Focusing on threshold selection of this important step for wavelet de-noising, proposing the threshold optimization de-noising approach based on the threshold of the critical point of difference;2.Adaptting wavelet threshold optimization methods and median filtering combination of remote sensing for marine oil spill removal SAR image noise. While effectively removing the impulse noise, removing Gaussian noise as well.